86 lines
3.0 KiB
Java
86 lines
3.0 KiB
Java
package eu.dnetlib.jobs;
|
|
|
|
import eu.dnetlib.featureextraction.FeatureTransformer;
|
|
import eu.dnetlib.support.ArgumentApplicationParser;
|
|
import org.apache.spark.SparkConf;
|
|
import org.apache.spark.ml.feature.CountVectorizerModel;
|
|
import org.apache.spark.sql.*;
|
|
import org.slf4j.Logger;
|
|
import org.slf4j.LoggerFactory;
|
|
|
|
import java.io.IOException;
|
|
import java.util.Optional;
|
|
|
|
public class SparkCreateVocabulary extends AbstractSparkJob{
|
|
|
|
final static int VOCAB_SIZE = 1<<18;
|
|
final static double MIN_DF = 0.1;
|
|
final static double MIN_TF = 1;
|
|
|
|
private static final Logger log = LoggerFactory.getLogger(SparkCreateVocabulary.class);
|
|
|
|
public SparkCreateVocabulary(ArgumentApplicationParser parser, SparkSession spark) {
|
|
super(parser, spark);
|
|
}
|
|
|
|
public static void main(String[] args) throws Exception {
|
|
|
|
ArgumentApplicationParser parser = new ArgumentApplicationParser(
|
|
readResource("/jobs/parameters/createVocabulary_parameters.json", SparkCreateVocabulary.class)
|
|
);
|
|
|
|
parser.parseArgument(args);
|
|
|
|
SparkConf conf = new SparkConf();
|
|
|
|
new SparkCreateVocabulary(
|
|
parser,
|
|
getSparkSession(conf)
|
|
).run();
|
|
}
|
|
|
|
@Override
|
|
public void run() throws IOException {
|
|
|
|
// read oozie parameters
|
|
final String workingPath = parser.get("workingPath");
|
|
final String vocabularyPath = parser.get("vocabularyPath");
|
|
final String vocabularyType = parser.get("vocabularyType"); //from file or from tokens
|
|
final double minDF = Optional
|
|
.ofNullable(parser.get("minDF"))
|
|
.map(Double::valueOf)
|
|
.orElse(MIN_DF);
|
|
final double minTF = Optional
|
|
.ofNullable(parser.get("minTF"))
|
|
.map(Double::valueOf)
|
|
.orElse(MIN_TF);
|
|
final int numPartitions = Optional
|
|
.ofNullable(parser.get("numPartitions"))
|
|
.map(Integer::valueOf)
|
|
.orElse(NUM_PARTITIONS);
|
|
final int vocabSize = Optional
|
|
.ofNullable(parser.get("vocabSize"))
|
|
.map(Integer::valueOf)
|
|
.orElse(VOCAB_SIZE);
|
|
|
|
log.info("workingPath: '{}'", workingPath);
|
|
log.info("vocabularyPath: '{}'", vocabularyPath);
|
|
log.info("vocabularyType: '{}'", vocabularyType);
|
|
log.info("minDF: '{}'", minDF);
|
|
log.info("minTF: '{}'", minTF);
|
|
log.info("vocabSize: '{}'", vocabSize);
|
|
|
|
Dataset<Row> inputTokensDS = spark.read().load(workingPath + "/tokens").repartition(numPartitions);
|
|
CountVectorizerModel vocabulary;
|
|
if (vocabularyType.equals("file")) {
|
|
vocabulary = FeatureTransformer.createVocabularyFromFile();
|
|
}
|
|
else {
|
|
vocabulary = FeatureTransformer.createVocabularyFromTokens(inputTokensDS, minDF, minTF, vocabSize);
|
|
}
|
|
|
|
vocabulary.write().overwrite().save(vocabularyPath);
|
|
}
|
|
|
|
}
|